A 2017 study by researchers in the United Kingdom has found that people who display various characteristics of autism but who are not diagnosed with autism are more at risk of attempting suicide than the general population.
Coventry University researchers found that people who exhibited higher levels of ASD traits were more likely to try to end their lives than people without the traits, the study found. Traits associated with autism spectrum disorders (ASD) include social and communication difficulties, excessive attention to detail, and the tendency to have narrow obsessive interests.
The reasons for attempted suicide ranged from feelings of exclusion from society to being a burden on friends and family, and because they experienced depression, researchers said. The study results were presented at the October 2017 meeting of the U.S. Department of Health & Human Services Interagency Autism Coordinating Committee.
As the researchers observed, scientists have already suggested a link between those with autism spectrum disorders (ASDs) and increased suicide risk. One study found that people with ASD are 7.55 times more likely to die by suicide, and a 2014 study, "Suicide in autism spectrum disorders," found that suicide occurs more frequently in high-functioning autism.
This study is different, though. Specifically, the researchers say it is the first study to suggest that those who have not been diagnosed with autism but had certain traits typical of autistic people were also more at risk of attempting suicide.
The Coventry University study involved 163 people aged between 18 and 30 who completed an online survey which asked a range of questions to establish the level at which they showed characteristics of autism, depression, and suicidal behavior.
According to the study, while recent research has shown that people with high autistic traits are more likely to attempt suicide, no studies have explored why.
"We found that people with high autistic traits were more likely to experience feelings that they do not belong in the world, are a burden on others, and depression, which may increase their likelihood of attempting suicide," the study stated. "These results suggest that promoting inclusion and independence in those with high autistic traits could help prevent people attempting suicide."
The researchers said a developing psychological model could be used to predict people who will exhibit suicidal behavior - and once further developed could make a huge difference in creating effective assessment, support, and treatment for people at risk of suicide.
The research was carried out by Mirabel Pelton and supervised by Dr. Sarah Cassidy.
"These are life and death issues," Pelton said. "Research to date has failed to understand the link between autistic traits and suicide but this study suggests that promoting social inclusion and independence could literally save lives."
Cassidy called the findings stunning.
"This is a shocking wake-up call for society, when a significant section of the population feels that they don't belong in the world," Cassidy said. "As a society we urgently must do more to value neurodiversity. This could ultimately save lives."
In her presentation to the IACC, Cassidy, who heads a group on mental health and autism, called the autism-suicide link an under-researched topic.
However, Cassidy said, through ongoing study, researchers have identified several key strategic areas in which future suicide and autism research should take place.
Identifying risk and protective factors, particularly developing new models to explain suicidality in autism; identifying the mechanisms underlying the increased risk; determining whether there are autism-specific risk factors - those and other areas of research will help researchers determine whether intervention and prevention strategies designed for the general population have to be changed to fit the needs of this group, Cassidy said.
The big question: Why mental health and suicidality in autism?
"There has been some recent research showing that the majority of autistic adults, 79 percent, meet criteria in their lifetime for at least one mental health condition," Cassidy said. "And we know from research in the general population that these mental health conditions are a significant risk factor for dying by suicide."
In 2014, Cassidy said, her group conducted a large-scale clinical study of suicidality in adults newly diagnosed with Asperger's syndrome and found what she called a really worryingly high rate of suicidal thoughts in that population.
"Sixty-six percent contemplated suicide," she said. "Thirty-five percent have planned or attempted suicide. And 31 percent have been diagnosed with depression."
And that 66 percent suicide ideation rate is significantly higher than the general population of 17 percent and patients with psychosis (59 percent), Cassidy said.
"It is even higher than the risk groups for suicidality," she said. "We also found that self-reported autistic traits and depression were significant risk factors for suicidality. And a much more recent and even larger scale study in Sweden showed that autistic adults were significantly more likely to die by suicide than the general population."
What was interesting and very important about that latter study, Cassidy said, was it indicated the risk factors for dying by suicide are very different from the general population.
"In particular, being female and being diagnosed with autism without a learning disability and depression risk factors," she said. "And being female ... is quite a key finding because in the general population, most people who die by suicide are males. It is a completely flipped-over risk profile."
As such, Cassidy said, most of the intervention and prevention strategies in the UK are aimed at men, but the results suggest something else again for autism.
While the scale of the problem is large and important, Cassidy said there is not enough research about why. For one thing, Cassidy said, there are questions whether the measures being used to assess suicidality in the general population are working in the same way or as effectively for people with autism.
"Difficulties such as alexithymia, which is a difficulty in articulating your internal or emotional experience, could lead to under- or over-reporting suicidality or depression," she said.
For example, Cassidy explained, many people with autism say the first thing that they are asked in a general practitioner appointment is how do you feel.
"That is a very difficult question for them to answer," she said. "They say I do not know how I feel. Most of the time that is a really impossible question to answer. They have mind difficulties. The ability to register or understand the impact of your death on another person. Or talking about suicidality to another person could affect the way that you talk about it."
Cassidy said behaviors such as social withdrawal or sleep problems that are indicators of depression in the general population overlap with characteristics of autism, but the challenge is to tease them apart.
"There might be unique aspects of suicidality in autism such as rigidity or reduced cognitive flexibility that impairs ability to weigh up alternatives that could really increase risk," she said.
Cassidy summarized ongoing research with which she is involved, particularly with respect to developing suicidality assessment tools for children with ASD or those with characteristics of autism.
The tools that do exist, Cassidy said, pose language difficulties.
"(One) has a question, something along the lines of do you have difficulties with sleeping, i.e., difficulty falling asleep, staying asleep, or sleeping too much," she said. "Things like that caused real issues. They just went with the first one or did not take into account the whole question. It led to really different responses and misinterpreted answers."
Respondents talked about difficulties in memory and time, Cassidy continued.
"If you asked in the last weeks, have you done any of the following, they found it really difficult to visualize that," she said. "They recommended maybe using a diary or a calendar."
And literal interpretation was a massive issue.
"Exactly how many thoughts," she said. "What about when I have accidentally overdosed but I did not mean to hurt myself or end my life? What do you mean by a plan?"
Assessment surveys also use insensitive language, Cassidy said, such as "commit suicide," which has a legal and moral connotation. Respondents are also worried about the purpose of assessments and wonder what will happen to them if they fill out a survey.
Cassidy then summarized three studies with which she is involved that relate to risk and protective factors, including the Coventry study that linked people with various characteristics of autism with higher risks of attempting suicide.
In that study, Cassidy said, researchers did not look at people with confirmed diagnoses of ASD.
"It is self-reported autistic traits in the general population that differ naturally in the general population," she said. "If you report a clinically significant level of autistic traits, you are more likely to have a diagnosis."
This study involved 163 general population young adults aged 18 to 30 years old. In the self-reporting population, Cassidy said, researchers found that self-reported autistic traits significantly predicted your likelihood of experiencing perceived burdensomeness and thwarted belongingness over and above age, gender, and depression.
"We also found that autistic traits significantly predicted suicidality through these constructs exactly as the model would predict," she said. "But these models and measures have been developed for the general population and we do not know how autistic people interpret the measures or how much the things that are included in the measures actually apply to autistic people."
In a second study, Cassidy said researchers formed a steering group of eight autistic adults who had experienced mental health difficulties and/or suicidality. The researchers developed an online survey from the topics and issues they explored with the steering group.
"We used the focus groups to identify themes, which may increase or decrease risks of experiencing mental health problems and/or suicidality and developed an online quantitative and qualitative survey to try and capture and measure these experiences," she said.
The key themes that emerged were isolation, not only social isolation such as lack of social connectedness or number of friendships, but also non-social isolation, Cassidy said.
"One person gave the example of not being able to engage in activities that they enjoyed on a day-to-day basis because they could not, for instance, get on the bus because of sensory overload, things like that and lots of other examples," she said.
The participants talked a lot about lack of belonging without being prompted, and lack of belonging in an autism unfriendly world, which led to thoughts of leaving, Cassidy said.
"It was completely kind of spontaneous," she said. "It was not prompted in any way. They just started talking about it. They talked about lack of opportunities in employment and education. For instance, having to leave a university or drop out of university because it was quite difficult for them."
People with autism can mask or camouflage their difficulties in an attempt to fit in, Cassidy said, which tends to lead to mental health problems and difficulties accessing help.
"They would present to services and they are so used to masking that they masked their distress and they do not come across as being that urgent," she said.
The focus group pinpointed problems with late diagnosis and misdiagnosis, too, Cassidy said.
"Quite a few of the people in the committee have been misdiagnosed with things such as borderline personality disorder before going on to get an autism diagnosis, but there are other examples," she said. "They also found that when they are diagnosed late, there was not any post-diagnostic support. They did not know where to go next."
The focus group also talked about childhood experiences, Cassidy said, about not being supported to have a positive identity or recognizing their strengths as well as their weaknesses.
From that exploration, Cassidy said a very large questionnaire was developed. She said 168 adults completed the entire survey, 67 males and 101 females and 108 control females.
In the autism group, Cassidy said, the suicide behavior scale was massively high - significantly higher than the recommended cutoff for psychiatric populations.
"It replicates what we found in a completely different environment and different method in our first study," she said. "But it is significantly higher than psychiatric groups."
There was a possible autism specific risk factor of camouflaging, Cassidy said.
"Camouflaging one's autism in an attempt to fit in in social situations was associated with significantly higher scores on this suicide behavior questionnaire just like the steering group describes," she said.
Finally, Cassidy talked about the group's psychological autopsy study, which aims to establish whether definite confirmed autism or possible autism diagnoses are overrepresented among people who died by suicide in the UK.
The psychological autopsy studies involved both analysis of coroners' inquest reports and interviews with friends and family of the person who died. The study is ongoing, and is being conducted in stages.
Cassidy said the researchers have looked at coroners' records for the period 2014 to 2015, recording a suicide as open where cause of death was undetermined, drug and alcohol related, or a narrative conclusion. They analyzed the records for any evidence of autism diagnosed or undiagnosed.
And what were the findings?
"So far we have looked at 219 coroner's inquest records, 150 which were ruled a likely suicide according to ICD-10 criteria, and 11 percent of these had evidence of autism, which were significantly higher than the general population rates of 1 percent," Cassidy said.
For the likely suicide, she said, the majority had no evidence, 9.3 percent had possible diagnosis, 1.3 had strong evidence, and one record contained a definite diagnosis.
"We know now, I think, that suicidality in autism does appear to be significantly higher than other psychiatric groups," she said. "However, because of difficulties in measurements and the types of data that we have had access to, it is unclear whether this is under- or over-estimated."
Still, she said, those with late diagnoses or who were undiagnosed without intellectual disability and adults without intellectual disability appear to be most at risk in a study from Sweden and in her group's study of late diagnosed adults with Asperger's syndrome as well as the coroners' study.
"They are all showing as the same thing," she said.
In addition, Cassidy said, there appears to be increased vulnerability to known risk factors for suicidality that overlap with the general population but that autistic people seem to be more at risk for, such as a reduced sense of belonging in the world, isolation, difficulties accessing support and treatment, as well as unemployment and comorbid mental health conditions.
"However, this is not the whole story because controlling for these differences and difficulties, there are still significant differences," she said. "And autistic people do still seem to be at significantly high risk than known psychiatric groups. You already have a risk. There is something else going on. It could be the measurement issue or it could be potential autism specific risk factors. And one that we potentially identified was camouflaging, but there may be others."
Richard Moore is the author of The New Bossism of the American Left and can be reached at www.rmmoore1.com.